What The PIXON Method Does

• simplest, i.e. smoothest, model for the image
• It must also be consistent with the data, i.e. have an acceptable chi-square fit.
• Being the simplest model, the derived image would be artifact free,
  i.e. there would be no spurious sources.

 A Pixon is a generalization of a pixel. It represents a single degree of
 freedom in the construction. Pixons are used to control the number of
 degrees of freedom by locally smoothing the image while still satisfying
the constraints imposed by the data.      (Metcalf, 1999)

The Pixon method is the slowest of the RHESSI reconstruction algorithms, mainly because it has to discover the best Pixon basis for the map (a computationally intensive process) as well as minimize the difference between the observed count rate profiles and the model countrate profiles.